Mangrove Phenology From Scale, Data and Species Perspectives

Land surface phenology derived from satellite observations provides an effective means of characterizing vegetation growth dynamics. Mangroves, although ecologically critical, remain understudied in phenology research due to geographic variation, inconsistent dataset selection, and limited understanding of species-specific patterns. In this study, we investigated mangrove phenology in the northern subtropical mangrove distribution area in Zhanjiang, China, from the perspectives of spatial scale, dataset resolution, and species differences. Using harmonic analysis to reconstruct the Enhanced Vegetation Index (EVI) time series, we assessed phenological trajectories at both regional and plot levels and compared Landsat 8 with Sentinel-2 observations. Results show that satellite-derived mangrove phenology exhibits a clear annual unimodal pattern across scales, with growth beginning in February, peaking in July, and continuing through December. Sentinel-2 outperformed Landsat 8 in capturing phenological signals, reflecting its superior spatial and temporal resolution. Pronounced interspecific differences were also detected: Aegiceras corniculatum and Bruguiera gymnorhiza exhibited continuous growth from January to December, whereas Avicennia marina and Rhizophora stylosa showed growth from March to December. Ground-based litterfall data revealed strong seasonality, with peaks in July-August. Except for A. corniculatum, litterfall exhibited a significant positive correlation with satellite-derived EVI growth patterns (p < 0.05), supporting the feasibility of species-level phenology detection from space. Given the increasing pressures on mangrove ecosystems from climate change and human activity, these findings highlight the importance and practicality of long-term phenology monitoring using multi-resolution satellite observations.